← Explore all mcp servers

SEMANTIC CODE TOOLS / Oraios

Serena MCP

Gives coding assistants symbol-aware navigation, editing and refactoring through language-server or JetBrains backends.

“Serena provides essential semantic code retrieval, editing, refactoring and debugging tools that are akin to an IDE's capabilities, operating at the symbol level and exploiting relational structure.”

01 / THE REASONING

Why this made the selection.

  • Offers project-aware semantic operations instead of relying entirely on text search and line offsets.
  • Tool registration marks edit-capable operations as destructive, exposing a useful permission distinction.

02 / THE REVIEW RECORD

What we actually inspected.

Source review has boundaries.
A clear record is more useful than a “safe” badge.

Material inspected

  • README.md at f9972a32c67359dd4c4a5ab597688bb10d21d1f3 (quick start, features, licence sections)
  • pyproject.toml at f9972a32c67359dd4c4a5ab597688bb10d21d1f3
  • src/serena/mcp.py at f9972a32c67359dd4c4a5ab597688bb10d21d1f3
  • LICENSE at f9972a32c67359dd4c4a5ab597688bb10d21d1f3

Our findings

  • The MCP entrypoint (src/serena/mcp.py) accepts the transports stdio, sse and streamable-http, and sets read_only_hint and destructive_hint on each tool from its can_edit() result.
  • The quick start installs serena-agent with uv, runs serena init (language-server backend by default, -b JetBrains for the plugin), then links to client-specific launch guidance; the README itself has no client configuration block.
  • The README lists the basic utilities search_for_pattern, replace_content, list_dir, find_file, read_file and execute_shell_command alongside the symbol-level retrieval, refactoring and editing tools.
  • At this commit the README announces Serena v2 with a REPL-based agent interface; pyproject.toml names the package serena-agent, version 2.0.1.dev0, licence GPL-3.0-or-later.

Not established by this review

  • Server startup, authentication and tool execution were not tested.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

Connect a server deliberately.

Upstream setup instructions ↗

DOCUMENTED COMMAND

uv tool install -p 3.13 serena-agent

Copying does not execute this command. It may retrieve a newer version than the reviewed source.

  1. Run the documented serena init step after installation; add -b JetBrains to use the JetBrains plugin backend instead of language servers.
  2. Choose the supported backend and follow the linked client configuration guide for the MCP launch command.

Before you start

  • uv and any language-server dependencies required by the project’s languages.

04 / FOLLOW THE EVIDENCE

The source trail.

Our notes are separate from the original resource.
Check upstream before adopting a new version.

  1. Official documentation ↗Checked

    https://github.com/oraios/serena/blob/f9972a32c67359dd4c4a5ab597688bb10d21d1f3/README.md

    Supports: summary, upstreamDescription, whySelected, bestFor, limitations, install, access, compatibility, transports

  2. https://github.com/oraios/serena/blob/f9972a32c67359dd4c4a5ab597688bb10d21d1f3/LICENSE

    Supports: license, limitations

  3. https://github.com/oraios/serena/blob/f9972a32c67359dd4c4a5ab597688bb10d21d1f3/pyproject.toml

    Supports: language, review

  4. https://github.com/oraios/serena/blob/f9972a32c67359dd4c4a5ab597688bb10d21d1f3/src/serena/mcp.py

    Supports: language, review

Evidence & Ask